Hi Ali,
Thanks for reply. I tried doing that by building a test script that build
gem5 and runs a program to see if floating point is working.m5 from
changeset 8178 passes while current gem5 fails. I am running into all sort
of compilation problems. The working

hg bisect -r
hg bisect -b
hg bisect -g 8178
hg bisect -c Test_m5_FP.sh

For example I get a lot this error:

 [     CXX] POWER_SE/cpu/base.cc -> .o
build/POWER_SE/cpu/base.cc: In member function 'virtual void
BaseCPU::startup()':
build/POWER_SE/cpu/base.cc:242:16: error: variable 'event' set but not used
[-Werror=unused-but-set-variable]

thanks,
Sameh

On Thu, Sep 6, 2012 at 10:07 AM, Ali Saidi <[email protected]> wrote:

> **
>
> Hi Sameh,
>
>
>
> If you could bisect the repository (see hg help bisect) and figure out
> what changeset causes the issues (it should take ~10 tries) that would help
> in debugging the problem.
>
>
>
> Thanks,
>
> Ali
>
>
>
> On 01.09.2012 19:12, Sameh Galal wrote:
>
> Hi everyone,
> I have used m5 couple of years back and managed to successfully get
> CFP2000 to work on powerpc SE target. When I tried to run the same
> benchmark on the latest version of gem5 all the floating point numbers were
> mysteriously zero. So I have written a small debugging program to
> demonstrate the issue. It asks the user for a real number and prints the
> square of that number. I cross compiled the program into powerpc executable
> and tried it on my 2 years old m5 files and the newest gem5. The program
> works correctly on the old m5 but displays the message "square of 0.0 is
> 0.0"  for nonzero input. I am attaching the 2 runs to the email and a zip
> file with the c file and powerpc executable. Any help will be greatly
> appreciated.
> thanks,
> Sameh
> *The code:*
>
> #include
>
> int main(void)
> {
>    double d, d_square;
>    int i, i_square;
>    printf("Please input an integer to square: ");
>    scanf("%d", &i);
>    i_square = i*i;
>    printf("square of %d is %d\n", i, i_square);
>    printf("Please input a real number to square: ");
>    scanf("%lf", &d);
>    d_square = d*d;
>    printf("square of %lf is %lf\n", d, d_square);
>    return 0;
> }
>  *compiling command:*
>  
> ~/crosstool/gcc-4.1.0-glibc-2.3.6/powerpc-750-linux-gnu/bin/powerpc-750-linux-gnu-gcc
> -static ./squarethis.c -o squarethis_ppc
>  *For the old M5 here is the log of the input and outputs:*
>  sameh06@drink-0:~/m5$ ./build/POWER_SE/m5.opt ./configs/example/se.py -c
> ~/squarethis/squarethis_ppc
> M5 Simulator System
> Copyright (c) 2001-2008
> The Regents of The University of Michigan
> All Rights Reserved
> M5 compiled Sep  1 2012 14:28:54
> M5 started Sep  1 2012 17:02:23
> M5 executing on drink-0
> command line: ./build/POWER_SE/m5.opt ./configs/example/se.py -c
> ~/squarethis/squarethis_ppc
> Global frequency set at 1000000000000 ticks per second
> 0: system.remote_gdb.listener: listening for remote gdb #0 on port 7000
> **** REAL SIMULATION ****
> info: Entering event queue @ 0.  Starting simulation...
> warn: allowing mmap of file @ fd 45172600. This will break if not
> /dev/zero.
> For more information see: http://www.m5sim.org/warn/3a2134f6
> warn: allowing mmap of file @ fd 45172600. This will break if not
> /dev/zero.
> For more information see: http://www.m5sim.org/warn/3a2134f6
> Please input an integer to square: 4
> square of 4 is 16
> Please input a real number to square: 5.3
> square of 5.300000 is 28.090000
> hack: be nice to actually delete the event here
> Exiting @ tick 10438500 because target called exit()
>  *For the gem5 the output is wrong :*
>  sameh06@drink-0:/~/gem5-stable$ ./build/POWER/gem5.opt
> ./configs/example/se.py -c ~/squarethis/squarethis_ppc
> gem5 Simulator System.  http://gem5.org
> gem5 is copyrighted software; use the --copyright option for details.
> gem5 compiled Sep  1 2012 16:48:31
> gem5 started Sep  1 2012 17:00:35
> gem5 executing on drink-0
> command line: ./build/POWER/gem5.opt ./configs/example/se.py -c
> ../FP-Genesis/FP-Gen/FP-Gen_ppc_study/squarethis/squarethis_ppc
> Global frequency set at 1000000000000 ticks per second
> warn: CoherentBus system.membus has no snooping ports attached!
> 0: system.remote_gdb.listener: listening for remote gdb #0 on port 7000
> **** REAL SIMULATION ****
> info: Entering event queue @ 0.  Starting simulation...
> Please input an integer to square: 4
> square of 4 is 16
> Please input a real number to square: 5.0
> square of 0.000000 is 0.000000
> hack: be nice to actually delete the event here
> Exiting @ tick 9296500 because target called exit()
>
>
>
>
> _______________________________________________
> gem5-users mailing list
> [email protected]
> http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users
>
_______________________________________________
gem5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/gem5-users

Reply via email to